Plasma-Sprayed Ti-6Al-4V Coatings in a Reactive Nitrogen Atmosphere up to 250 Kpa Vincent Guipont, Régine Molins, Michel Jeandin, G

Plasma-sprayed Ti-6Al-4V coatings in a reactive nitrogen atmosphere up to 250 kPa Vincent Guipont, Régine Molins, Michel Jeandin, G. Barbezat

To cite this version:

Vincent Guipont, Régine Molins, Michel Jeandin, G. Barbezat. Plasma-sprayed Ti-6Al-4V coatings in a reactive nitrogen atmosphere up to 250 kPa. International Thermal Spray Conference (ITSC 2002), Mar 2002, Essen, Germany. pp.247-252. hal-01480135

HAL Id: hal-01480135 https://hal-mines-paristech.archives-ouvertes.fr/hal-01480135 Submitted on 1 Mar 2017

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

Plasma-sprayed Ti-6Al-4V coatings in a reactive nitrogen atmosphere up to 250 kPa

V.GUIPONT, R.MOLINS, M.JEANDIN, Evry / F G.BARBEZAT, Wholen, CH

Abstract

Two spherical Ti-6Al-4V (25-45 µm and 45-75 µm) powders, were sprayed in a CAPS system ("Controlled Atmos- phere Plasma Spraying") operating in a coupled mode: High-Pressure Plasma Spraying (HPPS) and Reactive Plasma Spraying (RPS). Four different pressure settings up to 250 kPa with reactive nitrogen atmosphere were tested in order to assess the influence of chamber pressure and chamber atmosphere on the deposition of Ti-6Al- 4V coatings. The microstructures and phase compositions of the plasma sprayed Ti-6Al-4V coatings were studied using standard X-ray diffraction (XRD) and electron probe microanalysis (EPMA) with help of electron microscopy techniques (SEM and TEM). These established the pressure-assisted nitriding of the Ti-6Al-4V in the CAPS chamber with fine and coarse TiN precipitates embedded in a α-Ti matrix. A High-Pressure coupled with RPS enhanced the nitriding of the Ti-6Al-4V powder with a content of nitrogen which was all the higher because particle size was low.

1 Introduction sion-resistant coatings [9,10]. This mode could operate with a DC plasma gun operating with injected ni- “In-situ” chemical reactions between the melted mate- trogen reactive gas in a shrouding reactor [11] or with rial and its gaseous environment are generally inher- a controlled nitrogen atmosphere at different gas present to the thermal spray process. If these chemical re- sures in a chamber [12,13]. actions are promoted and controlled, this character- The aim of the present work was to achieve Ti-6Al-4V izes the reactive plasma spraying (RPS) mode. This nitrided coatings using high-pressure plasma spraying very powerful spraying technique allows to manufac- (HPPS) coupled with reactive plasma spraying in a re- ture composite materials, intermetallic alloys and rein- active nitrogen atmosphere, namely HPRPS (High forced or toughened ceramics [1]. Already with con- Pressure Reactive Plasma Spraying). The influence of ventional APS (atmospheric or air plasma spraying) chamber pressure and the influence of chamber at- mode, oxide compounds of metallic materials are syn- mosphere with some experiments using air atmos- thetized and located at the lamella boundary. If low ox- phere were studied. In addition, two spherical Ti-6Al- ide contents are generally targeted, a rather high oxide 4V (25-45 µm and 45-75 µm) powders were sprayed content might be interesting to enhance tribological in order to assess the influence of particle size on the properties for example. Thus, the APS mode can be resulting nitriding process.The formation of nitrided considered as the most common “reactive” plasma compounds was studied using standard X-ray diffrac- spraying mode to achieve multi-phased or composite tion (XRD) and electron probe microanalysis (EPMA) oxidized coatings. In case of titanium coatings sprayed with help of scanning and transmission electron mi- in the APS mode (with or without nitrogen in the croscopy techniques (SEM and TEM) in order to as- plasma gas), titanium particles react to form an oxi- sess the pressure-assisted nitriding of the Ti-6Al-4V. nitride of titanium and a solid solution of α-Ti that con- tains both nitrogen and oxygen [2]. 2 Materials and Processes More generally, RPS coatings can be achieved when bringing liquid, gaseous or solid precursors into con- 2.1 Powders tact with the sprayed material at a high temperature. To promote the chemical reaction, the reactants can Two different spherical, pre-alloyed and low-oxygen be the dissociated species or elements that form or content Ti-6Al-4V powders (PyroGenesis Inc., Can- interact with the plasma itself or the gaseous species ada) were used corresponding with two different parti- of the surrounding atmosphere of the process. More- cle size range: 25-45 µm and 45-75 µm. A SEM cross- over, if the reactive atmosphere is in contact with the section view of the powder (chemical etching with deposit that is held at a high temperature during spray- “Kroll” reagent) showed the martensitic structure of the ing, further reactions can occur after impact and solidi- Ti-6Al-4V alloy, Fig.1 . fication of the droplet. “In-situ” synthetized carbides or nitrides are the two types of RPS coatings that are al- 2.2 High-Pressure Reactive Plasma Spraying ways studied and corresponds to these various RPS (HPRPS) operating routes [3-7].In some works, pure titanium powder have been sprayed using RF plasma equip- Plasma spraying experiments were carried out using a ment with nitrogen content in the plasma gas [8]. In multi-process plasma equipment (CAPS, Sulzer- other works, titanium coatings containing titanium ni- Metco, Switzerland) with a F4-MB plasma gun. The trides were synthetized in a surrounding nitrogen at- CAPS system has an 18 m 3 chamber, which could mosphere. The latter mode is one of the most conven- operate in a controlled atmosphere of air, argon or ni- ient potential industrial process for wear and corro- trogen from 2 kPa to 350 kPa. The plasma spraying

parameters are given in Table 1 . The same Ar/He analysed with a 300kV EM430-T TEM (FEI, the Neth- plasma mixture was used for all HPRPS pressure set- erlands) coupled with EDS analysis. tings. Some experiments with air instead of nitrogen were performed and APS coating with Ar/H 2 plasma 3 Results and discussion was sprayed as a reference. The addition of N 2 in the plasma was studied in the HPRPS mode. Further ex- 3.1 HPRPS process in nitrogen atmosphere periments with a shorter spraying distance (75 mm) were done at 250kPa with a nitrogen atmosphere. In a previous work [14] using Ar/He in argon atmosphere up to 250 kPa, it was shown that the pressure concentrated the energy density within the HPPS plasma and improved the heat transfer between the Ti-6Al-4V particle plasma and the particle. When comparing an argon atmosphere with a nitrogen atmosphere, it was calcu- lated that the temperature of the plasma decreased more quickly along the plasma axis in a nitrogen atmosphere [15]. Nevertheless, Ar/He plasma rather than Ar/H 2 plasma (used in the conventional APS mode) showed the advantage to operate up to 250 kPa without overheating and damaging of the plasma nozzle for a nearly similar plasma effective electric power, Fig. 2 .

µ 20 m Fig. 1. Cross-section SEM image of a typical Ti- 6Al-4V particle (45-75 µµµm size range).

Table1. Plasma spray parameters

Reactive Plasma Spraying Atmosphere Nitrogen Pressure (kPa) 100 150 200 250 Plasma Ar:50/He:30, I=700A (l/min) Ar:50/He:30/N :2, I=650A 2 Ti-6Al-4V 45-75 µm or 25-45 µm Fig. 2. Plasma effective electric power (F4-MB gun) Distance 130 mm

Cooling gas argon The same plasma gas mixtures were kept whatever Atmospheric Plasma Spraying the applied pressure. However, when adding a low Atmosphere Air content of N 2 gas, it was necessary to reduce the arc Pressure (kPa) 100 100 250 current down to 650A to avoid any damage to the noz- Plasma Ar:47/H 2:10 Ar:50/He:30 zle due to the high enthalpy of N 2 plasma gas (nearly (l/min) I=650A I=700A similar to that of H 2). For this ternary plasma gas mix- Ti-6Al-4V 45-75 µm ture, a rather slight decreasing of the plasma effective Distance 130 mm electric power was obtained when pressure increased. Cooling gas air Note that HPRPS selected conditions were all with a lower plasma power than that for APS. Nevertheless, 2.3 Sample analysis this selected parameters with the cooling of the sub- strate allowed the explicit influence of surrounding re- As-sprayed samples were placed in a dry chamber active gas pressure to be assessed. This aspect was under vacuum to prevent any further oxidation. Phase a key issue of this study. analysis were performed using XRD D-500 goniome- ter (Siemens, Germany) with Co K α radiation (800W) 3.2 Influence of atmosphere on plasma- on the surface of the as-sprayed coatings (irradiated sprayed Ti-6Al-4V area = 10 mm²). EPMA quantitative analyses of cross- sections (with fittings for the Ti and N elements) were 3.2.1 Typical phase composition carried out using a SX 50 microprobe (Cameca, France) in the WDS. Cross-section observations of as Typical XRD diagrams of as-sprayed Ti-6Al-4V coat- –sprayed or etched coatings (“Kroll” reagent) were ings were selected to bring to the fore the titanium performed using a DSM 982 Gemini SEM (Zeiss, compounds (with help of the JCPDS cards) that could Germany) with back-scattered electron (BSE) detec- crystallize during the process, Fig. 3 . tor. Parallel foils of the coating were ion-thinned and

44-1294 (*) - Titanium - Ti 45-75mic./Ar-He/700A 87-0632 (C) - Osbornite, syn - TiN 45-75mic./Ar-He-N2/650A 76-0198 (C) - Titanium Nitride - Ti2N 77-2170 (C) - Titanium Oxide - TiO 25-45mic./Ar-He/700A 25-45mic./Ar-He-N2/650A 1500 1400 1200 1000 800 600 400 1000 200

TiN (200) peak area (a.u.) area peak (200) TiN 0 100 150 200 250 Nitrogen Pressure (kPa) D Counts 500 Fig.4. TiN (200) XRD peak area for different HPRPS C conditions.

The higher content of TiN was qualitatively determined for the HPRPS Ti-6Al-4V coating using the smaller particle size. In this case, pressure effect can be said B to be linear. This established clearly the gas pressure effect on the nitriding process in the HPRPS mode A 0 using nitrogen atmosphere. This effect was promoted by the use of a rather fine Ti-6Al-4V powder. Further 41 42 43 44 45 46 47 48 49 50 51 52 nitriding could be obtained, provided that 25-45 µm Ti- 2-Theta - Scale 6Al-4V powder is sprayed in the HPRPS mode using an Ar/He/N 2 plasma. Unfortunately, only one result in Fig. 3. XRD Diagrams: A: APS (100kPa, air atm., the RPS mode (100 kPa, N 2 atm.) was available. Ar/H 2 plasma), B: HPPS (250 kPa, air, Ar/He plas- Furthermore, pressure settings higher than 250 kPa ma), C: RPS (100 kPa, N 2 atm., Ar/He/N 2 plasma), D: could be interesting too. But, further investigation is HPRPS (250kPa, N 2 atm., Ar/He plasma) needed to define suitable plasma parameters at 350 kPa for example. If TiO and TiN peaks overlap, this could correspond to a titanium oxi-nitride phase. However, if Ti-6Al-4V 3.3 Nitrided Ti-6Al-4V coating microstructure coatings in air and nitrogen atmosphere are compared (A and D diagrams), a difference between these two 3.3.1 Nitrogen distribution within coatings compounds was clearly exhibited, i.e. TiO for APS and TiN for HPRPS. Spraying Ti-6Al-4V coatings in HPPS Using EPMA of HPRPS cross-sections (200 kPa, (Profile B) led to a titanium oxi-nitride. A high TiN con- Ar/He plasma, 25-45 µm specimen), the quantitative tent and a low α-Ti content were obtained in the distribution of the main constitutive elements (Ti, Al, V, HPRPS mode. In this case Ti 2N peak was also identi- N, O) was measured along a direction perpendicular fied. These results are in agreement with those ob- to the coating surface. Typical distribution of the nitro- tained for pure titanium [12] gen element within the lamellar HPRPS coatings could be obtained, Fig. 5 . 3.2.2 Pressure-assisted nitriding of Ti-6Al-4V The diagram could be divided in three different regions. These regions corresponded to three different On the basis of the value of the (200) peak area of the lamellae. Two of these lamellae were rich in nitrogen TiN phase (2 θ=50.1°), a qualitative assessment of the as well shown in the diagram (from 0 to 10 µm and nitriding level was obtained as a function of nitrogen from 18 to 24 µm) by mean of the nitrogen weight per- pressure in the CAPS chamber, Fig. 4 . This diagram centage which was rather high and constant in the featured the influence of the particle size and that of range of 11-15%. The third lamella (from 10 to 18 µm) nitrogen in the plasma gas. was a low-nitrogen content lamella (0-3%) that could be considered as a Ti-6Al-4V lamella.

Note that the specimen was free of oxygen except the to the different TiN contents obtained by XRD qualita- regions wherein pores were filled with mounting or- tive (see section 3.2.2) and because of the very differ- ganic resin. ent coating features. Using nitrogen in the plasma gas in the RPS mode (100 kPa, N 2) resulted in a well- wt%(N ) wt%(O ) wt%(Al) flattened and dense typical lamella structure, Fig 6-a- wt%(V ) wt%(Ti) 1. Nitrogen-rich regions were in dark grey while light grey regions corresponded to non-nitrided (or less- Nitrogen rich lamellae nitrided) lamellae using BSE SEM. Nitrided dark re- 15 100 gions were homogeneously distributed and were for around 50% of the coating. Using a high and reactive 14 90 13 pressure at 250 kPa (Ar/He plasma), large dark ni- 12 80 trided regions were observed, Fig 6-b-1. In this case, 11 a coating morphology with round particles was ob- 10 70 served. The high nitrogen content within HPRPS coat- 9 60 ings showed that the particles were melted, nitrided and probably partially solidified before impinging. Nev- 8 50 7 ertheless, the flattening ratio was sufficient to form a 6 40 plasma-sprayed coating. 5 At higher magnification, very different nitrided micro- 30 structures were exhibited, Figs. 6-a-2 and 6-b-2. The 4 Ti element % wt well-flattened RPS lamellae were with a small-sized N, O , N, O and V Al % wt 3 20 and homogeneous microstructure in the nitrided re- 2 10 1 gion. Some smooth and glassy regions were attributed 0 0 to the Ti-6Al-4V material (not represented in the fig- ure). This feature corresponded to an homogeneous 0 5 10 15 20 25 removal of Ti alloy or N-rich Ti alloy. This was con- step in microns firmed by TEM analysis, Fig. 7–a. Microcrystalline TiN particles were embedded and homogeneously distrib- Fig.5. EPMA profiles on HPRPS specimen uted in a pure or N-rich α-Ti matrix. On the other hand, the highly nitrided agglomerated HPRPS particles ex- This typical distribution of nitrogen within HPRPS coat- hibited an heterogeneous microstructure, as shown by ing showed that nitrogen was rather homogeneously selective chemical etching, Fig 6-b-2. Ti-N phases distributed on a micrometric scale within a nitrided la- (TiN and/or Ti 2N phases) with a typical inner dendritic mella. However, nitrided lamellae were randomly lo- microstructure and some dense layer at the particle cated near non- or poorly- nitrided lamellae. The re- periphery were observed [16]. The Ti-N compound sulting material can be considered as a Ti-6Al-4V was actually a corrosion resistant phase and could not based matrix composite with TiN-rich lamellar addition be etched by “Kroll” reagent. This revealed large areas (the size of which is the size of a splat). Despite the with both fine and coarse Ti-N dendrites. TEM view of pressure-assisted nitriding, this quantitative analysis this coating allowed to observe the corresponding mi- showed that the nitriding process did not affect homo- cro-scaled TiN particles and submicro-scaled TiN pre- geneously all the sprayed particles. Further work is cipitates, Fig. 7-b. These microcrystalline TiN particles needed to better control the TiN distribution within the were surrounded by an α-Ti alloy. Fine TiN precipi- HPRPS coatings. Argon cooling should probably be no tates corresponded to a high solidification rate while longer used to allow a better atmosphere homogenisa- coarser TiN particles were nitrided and partially solidi- tion. fied before impacted. As a confirmation of this, it was Furthermore, it was observed in nitrogen-rich regions observed that the dense peripherical layer was identi- that the weight percentages of all the Ti-6Al-4V consti- fied as sequence of coarse TiN particles. TiN or Ti2N tutive elements decreased, with a steep gradient for could not be easily detected using SEM of etched Al, i.e. from 6% down to 2-3%. No particular Ti, Al or V coatings and Ti 2N was not yet detected by TEM. increase in the rest of the coating that could re- Decreasing the spraying distance to 75 mm led to balance this chemical removal was detected using HPRPS coatings with a well-flattened lamella structure EPMA. This fact exhibited one distinctive characteris- and with a TiN content similar to that of the RPS coat- tic of Ti-6Al-4V reactive spraying compared to pure Ti. ing sprayed at 100kPa using Ar/He/N 2 plasma. This One may assume (but need confirmation) that inter- means that the nitriding process occurred mainly dur- mediate Ti-Al-V compounds could form then be ing the particle flight and could easily form high- evaporated during the HPRPS process. nitrided titanium compounds in the pressure–assisted process. Further investigation is needed to determine 3.3.2 Coating microstructure a spraying distance suitable for building-up well- flattened coatings with the highest content of nitrided SEM views of etched cross-section were achieved for titanium compounds. RPS and HPRPS coatings (25-45 µm particle size) using Ar/He/N 2 or Ar/He plasma gas mixture respec- tively, Fig. 6 . These coatings were selected according

Fig.6. SEM views, a: RPS (100 kPa, N2 atm., Ar/He/N2 plasma), b: HPRPS (250kPa, N2 atm., Ar/He plasma)

Fig.7. TEM views, a: RPS (100 kPa, N 2 atm., Ar/He/N 2 plasma), b: HPRPS (250kPa, N 2 atm., Ar/He plasma)

4 Conclusions HPRPS Ti-6Al-4V coatings at 250 kPa with small particle size powder exhibited the highest TiN content Pressure-assisted nitriding of plasma-sprayed Ti-6Al- with some Ti 2N phase. Pressure improved the heat 4V was obtained using CAPS up to 250kPa. Pressure transfer between the plasma and the particles and al- effect is a key parameter to carry out highly-nitrided Ti- lowed the use of a low enthalpy plasma to melt the based coatings because the nitrogen reacts mainly particles. In these conditions, the “in-situ” nitriding during the particle flight to form TiN or Ti 2N. With an process was so efficient that a typical dendritic growth increased pressure, it was possible to form nitrided of fine and coarse TiN particulates was observed lamellae within the coating even with a short spraying within sprayed lamella as confirmed by TEM. A lower distance. The use of nitrogen in the plasma gas was nitrogen content within HPRPS or RPS Ti-6Al-4V coat- not necessary to obtain a high nitrogen content. Using ings led to a lower TiN content without typical TiN a rather low enthalpy Ar-He plasma gas mixture cou- dendrites but with very fine microcrystallized TiN pled with a high-pressure of nitrogen in the chamber, and/or N-rich titanium solid solution and residual Ti-

6Al-4V martensitic phase. A promising application [6] P.V.Ananthapadmanabhan, P.R.Taylor, Tita- could be the tailoring of Ti-6Al-4V-based matrix com- nium carbide-iron composite coatings by reactive posites with TiN-rich lamellae containing titanium hard plasma spraying of ilmenite, Journal of Alloys and phases of different micro-scaled sizes that could im- Compounds 287 (1999), pp.121/25. prove wear properties. Work is in progress to charac- terize the TiN hard phase within HPRPS coatings us- [7] O.Al-sabouni, J.R.Nicholls, D.J.Stephenson, ing nano-indentation coupled with atomic force mi- Reactive plasma spraying of 80/20 nickel-chromium croscopy. powders, Journal of Materials Science Letter Some significant removal of Ti-Al-V was detected in 17(1998), pp.377/79. the N-rich lamellae of Ti-6Al-4V using EPMA and TEM analyses. Further work is needed to explain this par- [8] M.Fukumuto, S.Itoh, S.Itoh: Fabrication of ticular phenomenon and to discuss the role of Al and functionally gradient TiN Coating by reactive plasma V on the nitriding process. The particle size is another spraying, Proc.11th Int. Conf. on Surf. Modif. Technol., key parameter to manufacture nitrided Ti-based coat- 8-0 Sept. 1997, Paris, France, T.S. Sudarshan, et al. ings. If fragmentation into fine particles occurred dur- Eds. 1998, pp.306/18. ing the spraying process, it led to fine TiN round particulates. The direct use of ultra-fine or nanoscaled Ti [9] E.Lugscheider, H.Jungklaus, L.Zhao, based powder could be very interesting, especially in H.Reyman: Reactive plasma spraying of coatings con- case of co-spraying for the achievement of nitrides re- taining in situ synthesized titanium hard phases, Inter- inforced composite coatings for example. national Journal of Refractory Metals & Hard Materials 15(1997), pp.311/15. 5 Acknowledgements [10] T.Valente, F.P.Galliano, Corrosion resistance This study was done in the frame of the C2P (Center properties of reactive plasma-sprayed titanium com- for Plasma Processing/France) Club activity. Industrial posite coatings, Surface and Coatings Technology “members” of the Club are acknowledged for support 127(2000), pp.86/92. and helpful discussions. The authors thank Pyro- Genesis Inc. for kindly providing the Ti-6Al-4V pow- [11] E.Lugscheider, L.Zhao, A.Fischer: Reactive ders. Mrs N. De Dave, Mr G.Frot and Mr F.Borit are plasma spraying of titanium, Advanced Engineering also gratefully acknowledged for their work for sam- Materials 2(2000), Issue 5, pp.281/84. ples spraying and microstrutural analyses. [12] T.Valente, F.Carassiti, M.Suzuki, M.Tului: 6 References High pressure reactive plasma spray synthesis of titanium nitride based coatings, Surface Engineering [1] R.W.Smith: Reactive plasma spray forming for 16(2000) Issue 4, pp.339/43. advanced materials synthesis, Powder Metallurgy In- ternational 25, Issue 1, 1993, pp.9/16. [13] T.Bacci, L.Bertamini, F.Ferrari, F.P.Galliano, E.Galvanetto: Reactive plasma spraying of titanium in [2] C.Ponticaud, A.Grimaud, A.Denoirjean, nitrogen containing plasma gas, Materials Science P.Lefort, P.Fauchais: Plasma spraying of Ti particles - and Engineering A283(2000), pp.189/95. in flight reactivity - coating properties, in Thermal Spray 2001: New Surfaces for a New Millenium, Ed. [14] V. Guipont , M. Espanol, F. Borit, N. Llorca- C.C.Berndt et al., Pub. ASM Int., Materials Park, OH, Isern, M. Jeandin, K.A. Khor, P. Cheang: High- USA, 2001, pp.691/697. Pressure Plasma Spraying of Hydroxyapatite (HA) Powders, Accepted for publication in Materials Sci- [3] T.Eckardt, W.Malléner, D.Stöver: Reactive ence and Engineering A. Plasma Spraying of Silicon in Controlled Nitrogen At- mosphere, in Thermal Spray Industrial Applications, [15] M.Leylavergne, A.Vardelle, B.Dussoubs: Ed. C.C.Berndt et al., Pub. ASM Int., Materials Park, Comparison of plasma-sprayed coatings produced in OH, USA, 1994, pp.515/19. argon or in nitrogen atmosphere, in Thermal Spray: A united Forum for Scientific and Technological Ad- [4] Y.Tsunekawa, M.Okumiya, T.Kobayashi: Syn- vances, Ed. C.C.Berndt et al., Pub. ASM Int., Materials thesis of chromium nitride in situ composites by reac- Park, OH, USA, 1997, pp.459/65. tive plasma spraying with transferred arc, in Thermal Spraying: current status and future trends, Ed. [16] T.Bell, M.H.Sohi, J.R.Betz, A.Bloyce: Energy A.Ohmori, High Temp. Soc. Of Japan, 1995, beams in second generation surface engineering of pp.755/60. aluminium and titanium alloys, Scandinavian Journal of Metallurgy, 19(1990), pp.218/26. [5] R.W.Smith, Z.Z.Mutasim: Reactive plasma spraying of wear-resistant coatings, Journal of Ther- mal Spray Technology 1 (1992) Issue 1, pp.57/62.